Unwinding And The Medicaid Undercount: Millions Enrolled In Medicaid During The Pandemic Thought They Were Uninsured

Policy responses to the March 31, 2023, expiration of the Medicaid continuous coverage provision need to consider the difference between self-reported Medicaid participation on government surveys and administrative records of Medicaid enrollment. The difference between the two is known as the "Medicaid undercount." The size of the undercount increased substantially after the continuous coverage provision took effect in March 2020. Using longitudinal data from the Current Population Survey, we examined this change. We found that assuming that all beneficiaries who ever reported enrolling in Medicaid during the COVID-19 pandemic public health emergency remained enrolled through 2022 (as required by the continuous coverage provision) eliminated the worsening of the undercount. We estimated that nearly half of the 5.9 million people who we projected were likely to become uninsured after the provision expired, or "unwound," already reported that they were uninsured in the 2022 Current Population Survey. This finding suggests that the impact of ending the continuous coverage provision on the estimated uninsurance rate, based on self-reported survey data, may have been smaller than anticipated. It also means that efforts to address Medicaid unwinding should include people who likely remain eligible for Medicaid but believe that they are already uninsured.